Download PDF

The 84th Annual Conference of the Agricultural Economics Society
Edinburgh
29th to 31st March 2010
Modelling Climate Change Impacts on
Agriculture and Forestry with the Extended
LTEM (Lincoln Trade and Environment Model)
Caroline Saunders
William Kaye-Blake
Abstract
In the land-based sectors, agricultural production generally is a source of carbon, while
forestry may be thought to act as a sink. This paper focuses on new research examining the
interaction of the two. The core of the research is the Lincoln Trade and Environment model
(LTEM), a partial equilibrium model which links trade in NZ with the main trading countries
overseas, through to production and associated environmental consequences . This paper
reports on research expanding the model to include forestry from incorporating the
capabilities of the Global Forest Products Model (GFPM) into the LTEM and hence
producing an integrated model of agricultural and forestry land-uses for NZ and overseas. The
paper extends the environmental modelling capabilities of the LTEM to include the impacts of
climate change. The paper thereby reports on the development of a model of international
trade that encompasses major agricultural commodities and forestry, complete with linkages
and feedback with the environment and differentiated international markets. The paper then
presents results of scenarios around changes in consumer behaviour and production using the
new model.
Introduction
This paper focuses on extending the modelling capability to include forestry in an agricultural
partial equilibrium trade model. The core of the research is the Lincoln Trade and
Environment model (LTEM), a partial equilibrium model which links trade in NZ with the
main trading countries overseas, through to production and associated environmental
consequences. This paper discusses the issues, methodology and results from research
expanding the model to include forestry. This was done by incorporating the capabilities of
the Global Forest Products Model (GFPM) into the LTEM and hence producing an integrated
model of agricultural and forestry land-uses for NZ and overseas. The paper thereby reports
on the development of a model of international trade that encompasses major agricultural
commodities and forestry, complete with linkages and feedback with the environment and
differentiated international markets.
1
This paper initially considers the issues around gathering the necessary data and parameter
values to extend the LTEM to include the forestry sector, incorporate mitigation efforts and
technologies, and account for change in consumption related to climate change. It then details
the processes involved to include adding new variables to account for the commodities, new
parameters to account for specific features of forestry and the forestry-agriculture interactions,
and modifications to the equation structure to accommodate the new sector.
Finally, results from the updated model to investigate specific scenarios relating to climate
change, market reactions, mitigation efforts, and policies are presented. Scenarios were
developed based on various estimates of changes in agricultural production due to impacts of
climate change, based on the review of research in Kaye-Blake et al (2009). Some scenarios
also included potential changes in consumer behaviour in various markets in response to
climate change.
Main characteristics of LTEM
The LTEM is a partial equilibrium model of international trade in the agricultural sector, with
exogenous links to other industries, factor markets, and the macroeconomy. The LTEM is a
multi-country, multi-commodity model with a high degree of commodity disaggregation: the
dairy market is divided into five traded products and the oilseed complex is represented by
three commodities. The model quantifies price and quantity impacts on production,
consumption, and trade, and allows calculation of revenue and welfare impacts. The model
links through to the environment via production functions and then through to environmental
consequences. Currently, the model links through to groundwater nitrates, greenhouse gas
emissions, and energy.
The LTEM is a synthetic model since the parameters are adopted from the literature. The
symmetry condition holds for the supply and demand elasticities, therefore own- and crossprice elasticities are consistent. The model is used to quantify the price, supply, demand and
net trade effects of various policy changes. The policy parameters and/or variables and nonagricultural exogenous variables are listed in Table 1. The model is used to derive the
medium- to long-term policy impact in a comparative static fashion from the base year of
2004.
Table 1: Policy variables/parameters and non-agricultural exogenous variables
Policy VariablesDomestic Market
Policy VariablesBorder
Non-Agricultural Exogenous
Variables
Land set-aside
Import tariff
Gross domestic product
Production quota
Export subsidy
Country price index
Support/minimum price
Trade quota
Population
Producer market subsidy
In-quota tariff
Exchange rate
Producer input subsidies
Out-quota tariff
Producer direct payments
Export tax
Producer general services
Consumer market subsidy
2
The LTEM includes 22 commodities and 18 countries or regions. These are presented in
Tables 2 and 3. The dairy sector is modelled as five commodities, raw milk is defined as the
farm gate product and is then allocated to the liquid milk, butter, cheese, whole milk powder
or skim milk powder markets depending upon their relative prices, subject to physical
constraints. The meat sector is disaggregated into sheepmeat, beef and pig meat, and the
poultry sector (poultry meat and eggs) and wool are also modelled explicitly. There are eight
crop products (wheat, maize, sugar, other grains, rice, oilseeds, oil meals, oil) in the LTEM.
Table 2: Country coverage of the LTEM
Argentina
India
Russian Federation
Australia
Japan
South Africa
Brazil
Korea
Switzerland
Canada
Mexico
Turkey
China
New Zealand
United States
European Union (25)
Norway
Rest of World
World (data, market
clearing equations
Table 3: Commodity coverage of the LTEM
Wheat
Oilseed meals
Poultry meat
Cheese
Maize
Oils
Eggs
Whole milk powder
Other grains
Beef and Veal
Raw milk
Skim milk powder
Sugar (refined)
Pig meat
Liquid milk
Apples
Rice
Sheep meat
Butter
Kiwifruit
Oilseeds
Wool
A final general characteristic of the LTEM is that each commodity can appear in two different
forms. This allows researchers to model quality-differentiated products, such as two types of
wheat or two types of butter. The quality differentiation is linked to production methods. It is
thus capable of linking consumer preferences for specific production methods to the supply of
the products. The technique can be used for endogenising consumer demand for organically
grown food, genetically modified crops, or low-GHG emissions production.
Basically, the model works by simulating the commodity based world market clearing price
on the domestic quantities and prices, which may or may not be under the effect of policy
changes, in each country. Excess domestic supply or demand in each country spills over onto
the world market to determine world prices. The world market-clearing price is determined at
the level that equilibrates the total excess demand and supply of each commodity in the world
3
market. The LTEM is described in more detail in Cagatay and Saunders et al (2003) and
Saunders, et al and Cagatay et al (2003).
In general, there are six behavioural equations and one economic identity for each commodity
under each country in the LTEM framework. Therefore, there are seven endogenous variables
in the structural-form of the equation set for a commodity under each country 1 . There are four
exogenously determined variables, but the number of exogenous variables in the structuralform equation set for a commodity varies, based on the cross-price, cross-commodity
relationships. The behavioural equations are domestic supply, demand, stocks, domestic
producer and consumer price functions and a trade price equation. The economic identity is
the net trade equation which is equal to excess supply or demand in the domestic economy.
For some products the number of behavioural equations may change, as the total demand is
disaggregated into food, feed, processing industry demand, and are determined endogenously.
To simulate the impact of changing market conditions on production and thus the
environment, the factors affecting greenhouse gas emissions have been specified separately
and for the purpose of this study, emissions from beef and dairy cattle and sheep are taken
into account 2 . The principal determinants of gas from this source are livestock numbers, feed
intake and type per head (Lassey et al., 1992). Most animal waste decomposes aerobically on
pasture in New Zealand, resulting in relatively low levels of methane emissions from manure
management for this country (MfE, 2000). Lassey et al. (1992) also assesses emissions from
animal wastes, and from effluent processing plants such as abattoirs and dairy factories to be
of relatively minor importance.
The challenge of incorporating methane and nitrous oxide into the LTEM model is to produce
an equation (an environmental sub-module) which links all agricultural sources of these
greenhouse gases to domestic production, and measures the methane and nitrous oxide
emissions in physical terms. Therefore emission factors are crucial in this process, as well as
the effect of different production systems, domestic and border policies. The IPCC in its
guidelines produces default emission factors for different sources of gases, for a maximum of
eight regions of the world 3 . Greenhouse gases (GHG) are incorporated into the model through
the equation 32. In this equation GHG emissions from raw milk production is specified as a
function of applied nitrogen fertilizer (na) and number of animals (Naa) which are endogenous
to the model. The CH4 and N2O emission factors are implicit in the coefficients (ξ, ζ) and
values of these coefficients are provided by Clough and Sherlock (2001), equation 15. The
CH4 and N2O emissions from these sources are converted to their CO2 equivalents by
multiplying with their respective weights (21 and 310) to give CO2 equivalents 4 .
GHGamt = ξNa amt + ς (n at , Na at )
32
The calculation of coefficients for methane and nitrous oxide production from livestock
systems is based on the IPCC methodology for greenhouse gas inventories 5 . Methane and
1
The extended LTEM, including forestry, contains 4,767 equations, with each country having between 228 and
302 equations, depending on its primary sector. 2
In New Zealand, around 57 percent of methane emissions are from sheep and lambs, 27 percent from beef
cattle, and 17 percent from dairy cattle (MAF, 2001). 3
Naturally therefore, these values will vary considerably within each region, and New Zealand, as have many
other countries, has carried out in-depth research to provide more accurate emission factors. 4
The same equation is used to measure nation level emissions from beef and sheep also. 5
For details on these guidelines, see www.ipcc.org for ‘Revised 1996 IPCC Guidelines for National Greenhouse
Gas Inventories: Workbook’ 4
nitrous oxide are separated into their sources. Default emission factors provided by the IPCC
are used for the calculation of coefficients in most countries. In the case of nitrous oxide
production in New Zealand, the emission factors are based on recent research, and differ from
the default IPCC values. For the purposes of the model used in this study, coefficients
representing the total methane and nitrous oxide produced from all livestock sources, for each
animal type were calculated. Clough & Sherlock (2001) combined the emission factors for the
various sources into one coefficient for the production of nitrous oxide and one for the
production of methane per animal. A single coefficient for the nitrous oxide emitted from
nitrogen fertilizer was also calculated, constant across animals and countries.
Global Forest Products Model 6
The Global Forest Products Model (Buongiorno et al. 2003) is an economic model of the
global forest sector. The general principle of the model is that global markets optimise the
allocation of resources in the short-run (within one year). In the long-run resource allocation
is governed partly by market forces, as in trade, and also by external forces, such as waste
paper supply determined by environmental policy, tariffs by trade policy, and techniques of
production by technological progress.
The Global Forest Products Model deals with 180 countries (Appendix, Table 1), each of
which produces, consumes, imports, or exports at least one of 14 wood products (Appendix,
Table 2). The source of the base year production, consumption, trade, and price data for these
countries and products is the Food and Agriculture Organization online database FAOStat
(FAO 2008). These data are collected from individual country statistics, which it is
recognised, contain potential inaccuracies. However, the FAO is the only source of
internationally comparable country data. Furthermore, the calibration of the Global Forest
Products Model base year data (Buongiorno et al. 2001, Turner et al. 2005) addresses some of
the inaccuracies in the FAO data. While most trade data are left intact by the calibration
procedure, the production data are modified to ensure feasibility and consistency. For
example, consumption cannot be negative. Furthermore, the amounts of materials used in a
country and the amounts of products manufactured must be consistent with a priori
knowledge regarding the inputs needed per unit of output.
Because domestic price data are scarce for most countries, the market-clearing price in
countries that were net exporters of a product was assumed to be the world average export
unit value 7 . For net importers, the price was the world export price plus the freight cost and
import tariff for a particular country (Buongiorno et al. 2003, p. 75). Also needed for the base
year were country forest stock and forest area – from the Forest Resources Assessment 2005
(FAO 2005) – and GDP per capita – from the World Development Indicators database (World
Bank 2005).
From the base year, the Global Forest Products Model makes projections of forest resources,
and wood product prices and quantities to 2030. To make these projections the model requires
parameters describing the four main components of the wood-based sector: final demand, raw
material supply, manufacturing activities, and international trade. Demand for final products
and supply of raw materials are represented by econometric equations, which relate demand
6
For a detailed mathematical description of the Global Forest Products Model refer to Buongiorno et al. (2003)
Chapter 3. 7
The value of world exports divided by the volume of world exports for a product. 5
and supply volumes to product prices and gross domestic product. Manufacturing activities
are represented by input-output coefficients and manufacturing costs covering labour, energy
and capital. Transport cost depends on freight rates and import tariffs.
Each of the four components has a static and a dynamic element. The static part describes
each year’s competitive equilibrium – where the price of each product in each country is
solved so that consumption equals production plus imports minus exports. The dynamic
element is governed by endogenous changes – determined within the model – or exogenous
changes – determined outside of the model.
Demand for final products – fuelwood, other industrial roundwood, sawnwood, veneer and
plywood, particleboard, fibreboard, newsprint, printing and writing paper, and other paper and
paperboard – is represented by econometric equations (Buongiorno et al. 2003). These
equations relate the demand for each product to national income – measured by real gross
domestic product – and real product price – in U.S. dollars. The price and income elasticities
of demand – the percentage change in quantity demanded for a one percent change in product
price or country income – are in Table 3 in the Appendix. The Global Forest Products Model
determines real product price changes endogenously, that is simultaneously with the
quantities supplied, demanded, and traded. Country income changes – represented by the rate
of growth of real gross domestic product from World Bank (2008), OECD (2004) and EIA
(2004) reported in Turner et al. (2006) – are exogenous, reflecting assumptions regarding the
future economic growth of each country.
The supply, or harvest, of wood – fuelwood, industrial roundwood, and other industrial
roundwood 8 – is also represented by econometric equations (Turner et al. 2006a). These
equations relate wood supply to each country’s income per capita – measured by real gross
domestic product per capita –, forest stock, and wood price. The price, income per capita, and
forest stock elasticities – the percentage change in quantity supplied for a one percent change
in each explanatory variable – are in Table 4 in the Appendix. Wood price changes are
determined endogenously by the Global Forest Products Model so that they balance supply
and demand. The growth rates of country income per capita are exogenous, based on
assumptions regarding future economic and demographic growth (United Nations, 2005). In
the Global Forest Products Model they are meant to reflect the increase in wood supply due to
improvements in infrastructure and technology. Forest stock changes are determined
endogenously by the Global Forest Products Model, and reflect the harvest capacity of a
country.
The forest stock of a country is predicted with a growth-drain equation, where next year’s
stock equals the current stock plus the annual changes in forest stock due to forest area change
and to forest growth or decay on a given area, minus harvests. Stock change due to growth or
decay is a function of forest density – stock per unit area. Base year forest stock growth rates
are from FAO (2005).
Forest area change is a function of country income per capita, following the environmental
Kuznets curve for forestry. This suggests an increase in country income results in a declining
rate of deforestation at incomes below $9,000 per person. Above this income, as country
income grows there is an increasing rate of afforestation until an income of $21,000 per
person, after which the rate of afforestation declines until it is zero at $33,000 per person. The
theory upon which this representation of forest area change is based is sufficiently general to
8
Industrial roundwood and other industrial roundwood in the GFPM are not separated into softwoods and
hardwoods. 6
cover the economic situations of many countries, while being simple enough to implement
empirically with the scarce international data. Base-year forest area change rates are from
FAO (2005). The environmental Kuznets curve for each country is adjusted so that the
predicted forest area change rate for 2006 is equal to the observed.
The supply of waste paper is related to national income – measured by real gross domestic
product – and its real price – in U.S. dollars. Reflecting the availability of recovered paper,
there is an upper bound on waste paper supply, which is determined by a country's paper
consumption and recycling rate. This upper bound shifts over time due to endogenous
changes in paper consumption, and exogenous changes in the maximum recycling rate.
The assumed waste paper recovery rates were such that the world recovery rate would rise to
around 45 percent by 2030, from 39 percent in 2002 (Cesar 1995, Mabee and Pande 1997).
The supply of other fibre pulp – fibre from non-wood sources such as straw and bagasse – is
also related to national income and price.
The manufacture of wood products – sawnwood, veneer and plywood, particleboard,
fibreboard, mechanical and chemical pulp, newsprint, printing and writing paper, and other
paper and paperboard – is represented by input-output coefficients and associated
manufacturing costs. Input-output coefficients describe how raw materials are utilised in
production – the amount of input per unit of output – and differ among wood products and
countries. These data were estimated with the methods described in Buongiorno et al. (2001).
The manufacturing cost is the cost of the inputs – labour, energy, capital, etc. – not explicitly
recognized in the model. The manufacturing cost is an increasing function of the level of
production, described by an elasticity. For most manufacturing activities a one percent
increase in production results in a 0.10 percent increase in the cost of manufacture, apart from
the cost of wood and fibre inputs.
In the projections, manufacturing technology – represented by the input-output coefficients –
was held constant at its 2006 level, except for newsprint, printing and writing paper, and other
paper and paperboard. For these products the utilisation of waste paper in manufacture was
assumed to increase gradually, with a corresponding decrease in the amount of wood pulp
used, between 2006 and 2030. The estimated changes in waste paper utilisation were made by
extending historical trends and adjusting these trends according to expert opinion (Ince 1994,
Cesar 1995, Mabee and Pande 1997). It was also assumed that more wastepaper would be
utilised in regions where more wastepaper is recovered. The resulting increase in the
wastepaper utilisation rate was high – 0.70 percent per annum – in Asia and Oceania; medium
– 0.35 percent to 0.50 percent per annum – in Europe, South America, former USSR, and
North America; and low – 0.20 percent per annum – in Africa. For countries with already low
levels of wood pulp utilisation the anticipated increase in waste paper utilisation was slower.
The Global Forest Products Model predicts trade flow volumes – between each country and
the world market – for all wood products, except other industrial roundwood. Predicted trade
flows are influenced by the cost of transportation, which includes the cost of freight and
import tariffs. Freight costs are those reported in Turner and Buongiorno (2001). Import tariff
data for 2006 were from the APEC 9 and UNCTAD TRAINS 10 databases (Turner et al.
2006b).
9
www.apectariff.org www.unctad.org/trains 10
7
Using the GPFM to expand the LTEM
As the above descriptions indicate, the GFPM and LTEM are structurally similar, although
they focus on different commodities and countries. The similarities between the two models
made them ideal for combining into a single model. The LTEM was chosen as the framework
for the combined model, and it was expanded using material from the GFPM. This chapter
discusses how that was done.
The LTEM contains several different equation structures for different commodities. Field
crops, for example, are treated differently from livestock production. For the present work, the
structure of the livestock equations was used, for two reasons. First, forestry is most likely to
compete with pastoral agriculture for land use. Secondly, both livestock and forestry have
current production levels that depend on available stock, and thus on prior production levels.
The general form of the forestry equations are shown in equations 43 and 44:
qdi , ft = β 0 pcit β1 pinct β2 popt β3 ∏ pc jt j ;
β
β1 < 0 , β 2 > 0 , β3 > 0 , β j > 0
43
j
qs f = θ f 0 pp f
θ ff
∏ pp
θ fj
j
;
θ ff > 0 , θ fj < 0
44
k
The demand for forestry products is a function of the price, as well as personal income,
population, and the prices of other products. The supply of forestry products is a function of
its own price, the prices of other forestry products, and the prices of agricultural products. The
responsiveness is a function of the elasticities, given as β and θ.
The GFPM is highly disaggregated by country and forestry product. For the purposes of
examining the impact of climate change on New Zealand, a more aggregated description of
the forestry sector was sufficient. The number of forestry products was reduced from 14 to
four: firewood, roundwood, panelwood, and paper and pulp. Table 4 provides the mapping
from the GFPM products to the extended LTEM products.
Table 4: Mapping of GFPM Products to LTEM Commodities
8
GFPM Product
LTEM Product (Code)
Fuelwood and charcoal
Firewood (FWD)
Industrial roundwood
Roundwood (RWD)
Other industrial roundwood
Roundwood (RWD)
Sawnwood
Panelwood (PWD)
Plywood
Panelwood (PWD)
Particleboard
Panelwood (PWD)
Fibreboard
Panelwood (PWD)
Mechanical pulp
Paper and pulp (PPP)
Chemical pulp
Paper and pulp (PPP)
Other fibre pulp
Paper and pulp (PPP)
Waste paper
Paper and pulp (PPP)
Newsprint
Paper and pulp (PPP)
Printing and writing paper
Paper and pulp (PPP)
Other paper and paperboard
Paper and pulp (PPP)
Quantities produced were calculated for each of the four aggregate products. Prices were
calculated as weighted averages of the prices of the constituent products in the GFPM. These
quantities and prices were then used for the equations described above.
Some countries are present in both the GFPM and the LTEM. Data from these countries were
added to the database for the LTEM. Other countries in the GFPM are either part of regions in
the LTEM, or included in the Rest of the World (ROW). For these countries, production data
were summed and transferred to the LTEM. Price data were aggregated with weighted
averages and then transferred to the LTEM database.
Finally, the data on trade policies in the GFPM was also incorporated into the LTEM. The
GFPM uses producer subsidy equivalents (PSEs) to model the impact of trade policies, and
these can be incorporated multiplicatively into supply equations in the LTEM. The PSEs are
also used to calculate consumer subsidy equivalents (CSEs) for the LTEM, to maintain the
domestic balance between the producer and consumer prices in the model.
Many of the equations in the LTEM use elasticities to model the reaction of a dependent
variable to changes in an independent variable. For example, price elasticities of demand are
used to model the change in consumption that results from a given change in price. The
GFPM similarly contains elasticities for supply, demand, and other equations. Some of these
elasticities translated directly into model inputs for the LTEM. Other elasticities required
calculation, and these were generally calculated by finding weighted averages of elasticities
across the products and countries that were be aggregated. Finally, some of the calculated
elasticities were adjusted before they were included in the LTEM. The supply and demand
equations in the LTEM are somewhat different from the input-output structure of
manufactured forestry products in the GFPM. Price elasticities of demand in the LTEM were
thus constrained to be less than -0.20 (that is, greater than an absolute value of 0.20), which
allowed the model sufficiently flexibility to find solutions to the climate change scenarios.
9
Cross-elasticities are also used in the LTEM to model the interaction between commodities.
In the original LTEM, for example, the supply of beef is influenced by the price of sheep and
milk. The responsiveness is a function of the size of the cross-elasticity.
Data were sought from the literature on land use change, particularly in New Zealand, in order
to understand the interaction between agricultural and forestry products. The researchers also
consulted with colleagues at Motu Economic and Public Policy Research in Wellington and
the office of the Parliamentary Commissioner for the Environment. The results of empirical
work in New Zealand conducted by Suzi Kerr and Jo Hendy suggest that the elasticity of
supply of forestry products with respect to the price of agricultural commodities is quite low,
and the same is true for the supply of agricultural commodities with respect to the price of
forestry products 11 . In order to incorporate these interactions, therefore, low cross-price
elasticities were included in the extended LTEM.
The material from the GFPM was incorporated into the LTEM by using the existing LTEM
equation structure, aggregating data from the GFPM and adding them to the LTEM, and
including supply and demand elasticities, and cross-elasticities in the equations.
Scenario Descriptions
The expanded LTEM was used to model several different scenarios. These scenarios were
designed around the two goals of the research: developing a new model to assess climate
change, and investigating the impact of climate change and reactions to it. To help develop
the model, scenarios were designed to test the expanded model, to investigate how it reacts to
different types of inputs, and to identify areas of future work to improve results and their
applicability. To provide information about potential impacts of climate change, inputs from
the literature on climate change, carbon emissions, and evolving market trends were
incorporated into scenarios.
The scenarios were designed around four dimensions. The first dimension was the presence of
climate change. Some scenarios included impacts of climate change estimated from the IPCC
scenario A2. However, there are significant uncertainties around future climate change and
the impacts on agricultural productivity. Some scenarios without climate change impacts were
thus included. This approach allows the results to be used more widely for understanding
potential impacts of policy and market trends, holding the level of climate change constant.
To model climate change effects, the productivity impacts described in Kaye-Blake, et al.
(2009) were used to modify the supply shift parameters, which is shfqs in equations 38 and 39.
The second dimension considered in the modelling was the extent of policies to curb
greenhouse gas emissions. A number of policy tools have been discussed in the literature,
including carbon taxes and cap-and-trade policies. These policies all have the impact of
placing a direct or implicit price on carbon. They can all be modelled similarly, that is, as an
increase in the cost of production that is proportional to the amount of GHG emissions. They
were therefore modelled as changes to the supply shift parameter in equations 38 and 39. The
impact on productivity was calculated based on emissions from beef, sheep, and dairy in the
different countries and a price of US$25 per tonne of CO2 equivalents. This productivity
impact was then used to calculate a new supply shift parameter. The policies were further
divided into two possibilities. One possibility is that all Annex 1 countries include agriculture
in greenhouse gas emissions policies, and Non-Annex 1 countries are exempt. The second
possibility is that New Zealand includes its agricultural sector in its ETS, but no other country
11
Pers. Comm., Jo Hendy, 6 November 2009, and Wei Zhang, 5 November 2009. 10
follows suit. Both of these possibilities have been modelled. For all policies, forestry products
were modelled as carbon-neutral, and therefore not affected directly by GHG policies.
The third dimension that formed part of the scenario development was the use of mitigation
technologies. As discussed in Kaye-Blake, et al. (2009), there are techniques and technologies
with the potential to reduce greenhouse gas emissions from agriculture. If these technologies
are implemented, there are two impacts. They reduce the potential liability from GHG
policies, reducing the added costs that the primary sector would incur from such policies. In
addition, they reduce the amount of GHG emitted per unit of production. In the present
research, mitigation technologies were modelled alongside GHG policies, to investigate the
joint impacts of technological improvements and price signals. Around one-half of New
Zealand production was modelled as having no reductions in emissions, while the other half
was modelled was having a 30 per cent reduction. This level of reduction is based on the
scientific research discussed in Kaye-Blake, et al. (2009), and represents some of the highest
levels of reduction. This mitigation level may therefore represent the potential of current
research, rather than mitigation that is actually achievable on-farm.
For the scenarios presented in this paper, the split-commodity capability of the model was
employed. The production in every country was divided evenly between standard production
and low-emissions production. Between the two methods of production, each commodity was
highly substitutable to avoid constraining production of one type or the other. The standard
production method produced the current (2004) level of greenhouse gases. The low-emissions
method produced 30 per cent less GHG emissions per unit of production. This difference was
modelled by adjusting the supply shift parameters so that the low-emissions product had a 30
per cent lower shift than the conventional product.
The final dimension considered in the scenarios was consumer demand for lower-emissions
methods of production. In some scenarios, low-emissions product did not attract a price
premium and were not preferred by consumer. Other scenarios included a 10 per cent shift in
demand for low-emissions products, representing a price premium that consumers would be
willing to pay for production method with lower GHG emissions. The premium was modelled
with the demand shift parameter, shfqd in equation 42. For the low-emissions product, the
parameter was set at 1.10, while it was set at 1.00 for the standard product.
Altogether, the results of 15 scenarios are included in this paper. They are summarised in
Table 5, which includes the scenario code and ticks indicating which element or elements
were included in the scenario. As the table indicates, GHG policies could be enacted either by
all Annex 1 countries or just New Zealand, and mitigation technologies could appear
alongside GHG policies.
Table 5: Scenarios modelled
-- GHG policies -Scenario
code
Climate
change
All Annex
1
01
9
02
9
03
9
04
9
NZ only
Mitigation
technologies
9
9
11
Low-emissions
demand
05
9
9
9
06
9
07
9
08
9
9
09
9
9
9
9
9
10
9
12
13
9
15
9
9
9
9
17
19
9
9
12
9
9
9
9
9
9
9
9
Results
The model was used to investigate the impact of several different future scenarios on the
agricultural and forestry sectors in New Zealand. The results of modelling these different
scenarios are presented below. Two summary measures are used to describe the impact of
each scenario. The first is a financial measure: the net change in producer returns. Producer
returns indicate the total revenue earned by a sector, and are calculated by multiplying the
amount of a commodity produced in New Zealand by its price. The second measure is the
change in greenhouse gas emissions. The model focused on the production of methane and
nitrous oxide from animal production, as well as total greenhouse gas emissions from
agriculture. The change in emissions from animals is based on the number of animals
produced and the uptake of emissions-reducing techniques and technologies.
Scenarios with climate change
The first set of scenario results are based on the climate change scenarios developed for IPCC
research. The trade model was modified to reproduce the productivity impacts expected under
climate change scenario A2. These impacts affected both agricultural and forestry
commodities, and have been estimated for several regions and many specific countries,
including New Zealand. The productivity impacts were then placed alongside other potential
changes in the agricultural and forestry sectors, and the net results calculated.
The results from these scenarios are presented in two tables. Table 6 presents the percentage
changes in producer returns expected under the different scenarios. The producer returns are
presented for all New Zealand agriculture, and then for the separate industries of beef,
sheepmeat, and dairy. The final column provides the impact on producer returns for
roundwood production.
The first scenario examined the expected impacts on New Zealand of worldwide climate
change under IPCC climate scenario A2, and is scenario code 03. With climate change,
production in some regions and countries declines, while in others, production increases. New
Zealand productivity declines, but not as much as in some other countries. Reduced quantities
of commodities also lead to higher prices. The net result is that a scenario including only
climate change and no policy or market impacts produces an increase in producer revenues in
the New Zealand primary sector. Beef revenues decline slightly, as a result of higher impacts
on dryland pastures in New Zealand and productivity gains overseas, such as in the United
States. Sheepmeat and dairy revenues increase, a combination of domestic productivity
impacts, overseas climate changes, and New Zealand’s contribution to international trade of
these commodities. Forestry production also increases, as a result of increased productivity.
The second scenario in Table 6 includes both the climate change impacts as well as
implementation of GHG policies in all Annex 1 countries at US$25 per tonne. The policies
are modelled in the LTEM as affecting the cost of production and thus reducing the
productivity of farmers: increased inputs are required to produce the same level of outputs. As
a result, greenhouse gas policies reinforce the impacts of climate change. Production becomes
more expensive, commodity prices increase, and the primary sector producer revenues
increase. Producer returns in forestry are constant. Forestry products are modelled as ‘carbon
neutral’ and thus not affected directly by GHG policies. The indirect impacts from land use
change are not large enough to affect overall producer returns.
13
Table 6: Percentage changes in New Zealand producer returns, climate
change scenario A2
Scenario (code)
Climate change only (03)
All
agriculture Beef Sheepmeat Dairy Roundwood
14.6
-0.9
18.2
21.5
9.2
With worldwide GHG policy (04)
With worldwide GHG policy and
mitigation (05)
31.0
2.2
32.2
55.2
9.2
28.3
1.7
29.8
49.6
9.2
With NZ-only ETS (08)
With NZ-only ETS and mitigation (09)
7.6
8.6
-8.0
-7.1
13.5
14.2
18.3
18.7
9.1
9.1
The second scenario in Table 6 includes both the climate change impacts as well as
implementation of GHG policies in all Annex 1 countries at US$25 per tonne. The policies
are modelled in the LTEM as affecting the cost of production and thus reducing the
productivity of farmers: increased inputs are required to produce the same level of outputs. As
a result, greenhouse gas policies reinforce the impacts of climate change. Production becomes
more expensive, commodity prices increase, and the primary sector producer revenues
increase. Producer returns in forestry are constant. Forestry products are modelled as ‘carbon
neutral’ and thus not affected directly by GHG policies. The indirect impacts from land use
change are not large enough to affect overall producer returns.
The third scenario in Table 6 shows the impact of including mitigation efforts in the model
alongside worldwide GHG policies and climate change. Mitigation reduces some of the
impacts of GHG policies: producers become more ‘carbon efficient’ and therefore have lower
costs associated with the policies. As a result, their productivity relative to other producers is
increased and price are lower on average. For agricultural products, the net result is a decrease
in producer returns relative to a scenario with no mitigation, but the returns are higher than in
a scenario with no GHG policies at all. Roundwood again shows no change.
The fourth and fifth scenarios in Table 6 indicate the impacts on New Zealand from global
climate change, but only a domestic GHG policy, such as the ETS. Other Annex 1 countries,
in these two scenarios, exempt their agricultural sectors from GHG policies. In addition, the
fifth scenario also includes mitigation technologies, which have economic impacts only in
New Zealand. Under these conditions, New Zealand does gain in relation to the baseline, as a
result of higher prices and lower worldwide production brought about by climate change.
However, relative to other climate change scenarios, New Zealand primary sector producers
have lower revenues. The difference relative to the scenario with no GHG policies at all is a
seven per cent reduction in producer returns across agriculture (forestry is essentially
unchanged, although results suggest downward pressure on the industry). With mitigation,
agriculture is able to regain one percentage point of the difference, but is still below the nopolicies scenario. Of the livestock sectors, dairy is the least affected.
The model also allowed calculation of the impact on GHG emissions from agriculture and
forestry of the different scenarios. The results are presented in Table 7. The scenarios are the
same as those discussed with the previous table.
14
Table 7: Percentage changes in New Zealand methane and nitrous oxide emissions,
climate change scenario A2
All
livestock
0.1
Beef
-9.8
With worldwide GHG policy (04)
With worldwide GHG policy and mitigation (05)
-1.3
-14.3
-18.6
-28.4
-1.2
-14.6
10.2
-4.6
With NZ-only ETS (08)
With NZ-only ETS and mitigation (09)
-7.4
-18.8
-16.6
-26.8
-7.0
-18.9
-1.6
-13.4
Climate change only (03)
Sheepmeat Dairy
0.9
5.8
Climate change is expected to reduce agricultural production in general, and regional
variation is also expected. The impact on New Zealand is partially bio-physical, that is, the
amount of production that could be sustained given soils, climate, etc. The impact is also
partially a result of changes to production that flow through to international markets. If
production falls overseas for commodities of which New Zealand is a major supplier, then the
country is likely to see a large impact. If other suppliers of a commodity are not significantly
affected, or even see increases in production (such as are predicted for some regions in some
climate change projections), then New Zealand production could even decline.
The results presented in Table 7 indicate that these different pressures on production and
markets will have uneven impacts across New Zealand agriculture. For example, climate
change scenario A2, when modelled with the LTEM, led to increases in dairy production and
thus increased GHG emissions, nearly constant production in the sheep sector, and decreases
in beef production with accompanying falls in emissions.
The unevenness of the impacts is exacerbated by worldwide GHG policies. Implementation of
policies leads to general decreases in New Zealand emissions from livestock. However, the
beef sector reduces emissions by nearly 20 per cent, while the dairy sector actually increases
its emissions by over ten per cent. If mitigation technologies are implemented worldwide
alongside carbon charges and climate change, then New Zealand beef and sheep producers
have large decreases in emissions, while dairying has smaller reductions.
The general pattern is repeated in the scenarios in which only New Zealand implements GHG
policies. Emissions fall, mirroring the fall in producer returns discussed above, but fall the
most in the beef sector and least in dairy. Mitigation technologies reduce emissions even
more, with the livestock sectors showing an overall decrease of nearly 20 per cent. Once
again, these decreases are achieved unevenly across the sectors.
In all of the above scenarios, no consumer reaction was included. As discussed above, other
scenarios modelled with the expanded LTEM also included a ten per cent demand premium
for low-emissions products. This premium was applied in several cases, and the results are
presented in Table 8.
15
Table 8: Percentage changes in New Zealand producer returns with demand for
lower-emissions products
All
agriculture
Beef
Demand plus climate change (13)
51.1
14.0
52.0
72.7
20.9
Demand plus climate change, GHG
policies, and mitigation (15)
70.0
17.3
67.0
113.2
20.9
Demand plus climate change, NZ-only
ETS and mitigation (19)
43.7
7.1
47.6
69.7
20.8
Sheepmeat Dairy Roundwood
The first possibility considered was that the response to climate change would be left to the
market. If consumers were concerned about their impacts on GHG emissions from
agricultural production, then they could pay more for low-emissions production methods. In
combination with climate change, the demand premium led to an overall increase in
agricultural producer returns of 51.1 per cent. The dairy sector saw the largest increase, while
the beef sector had the smallest. Returns for roundwood production also increased, by over 20
per cent. Simply put, increased demand overseas for desirable primary products created
significant increases primary sector revenues.
The impact of GHG policies were also considered, both policies implemented by all Annex 1
countries and a New Zealand-only policy. In both cases, primary sector producer returns
increased, and they increased more than in any of the scenarios in which consumer responses
were not considered. When the GHG policies are implemented worldwide, New Zealand
gains significantly from the decreased productivity and increased demand. If New Zealand is
alone in implementing such policies, then the gains are not as large. As with the earlier
scenarios, the results are spread unevenly across the three livestock sectors.
The changes to GHG emissions in these scenarios were also calculated, and they are presented
in Table 8. For all scenarios, the combination of climate change and demand for loweremissions products leads to a general reduction in agricultural GHG emissions. However,
emissions from dairy tend to increase, except in the case in which New Zealand is the only
country implementing a GHG policy. Emissions from beef decline significantly, and
emissions from the sheep sector also decrease.
Comparing this to Table 7 indicates an interesting result. With no policy in place regarding
emissions or mitigation, consumer demand for lower-emissions production leads to lower
emissions. Emissions are reduced by about six per cent overall, while climate change alone
did not reduce emissions. However, the reduction in emissions is actually lower in the other
two scenarios than in their counterparts in Table 7. The reason for this result is that the
increase in demand for lower-emissions products leads to a net increase in production and
thus in emissions.
16
Table 9: Percentage changes in New Zealand methane and nitrous oxide emissions,
with demand for lower-emissions products
Demand plus climate change (13)
Demand plus climate change, GHG policies,
and mitigation (15)
Demand plus climate change, NZ-only ETS
and mitigation (19)
All
livestock
-5.8
Beef
-25.1
Sheepmeat
-5.1
Dairy
6.2
-6.8
-31.6
-7.1
9.9
-11.9
-30.1
-11.6
-0.2
Scenarios without climate change
Another set of scenarios removed the impacts of climate change. These results indicate the
impacts of GHG policies, mitigation, and consumer reactions, without the additional impacts
of climate change.
Table 10 provides the changes to New Zealand producer returns under four different
scenarios. The first scenario is the implementation of GHG policies in all Annex 1 countries.
These policies increase the cost of producing agricultural products, reducing production. The
result is an increase in market prices. The net impact on New Zealand agriculture is an
increase in producer returns. For forestry, GHG policies are modelled as neutral so there is no
impact on forestry returns.
Table 10: Percentage changes in New Zealand producer returns, climate change
impacts removed
With worldwide GHG policy (01)
With worldwide GHG policy and
mitigation (02)
With NZ-only ETS (06)
With NZ-only ETS and mitigation
(07)
All
agriculture
13.5
Beef
3.5
11.2
2.9
9.5
22.1
0.0
-5.9
-6.5
-3.8
-2.6
-0.1
-5.1
-5.6
-3.3
-2.2
-0.1
Sheepmeat Dairy Roundwood
11.5
26.6
0.0
The second scenario in Table 10 combines Annex 1 GHG policies with mitigation
technologies. This combination leads to increased producer returns in agriculture, with large
gains for dairy and lower returns for beef and sheepmeat. The increases are somewhat lower
than in the previous scenario, as mitigation technologies reduce the costs of GHG policies.
The next two scenarios examine the impacts of a GHG policy implemented only in New
Zealand. In both scenarios, the producer returns in New Zealand are reduced. Returns for beef
fall the most, while returns in the dairy industry fall least. The forestry sector remains
essentially unchanged, with a margin impact from interactions with other commodities. With
mitigation technologies, the reduction in producer returns is lessened as a result of lower costs
for GHG emissions.
The impact on GHG emissions were also calculated for these scenarios, and reported in Table
11. With GHG policies in all Annex 1 countries, emissions are somewhat reduced overall, but
17
the impacts are uneven. Emissions from dairy increase, as New Zealand increases its
production to replace reduced production overseas. Emissions from beef and sheep production
in New Zealand decline. With mitigation included alongside the GHG policies, emissions fall
for all the commodities.
Table 11: Percentage changes in New Zealand methane and nitrous oxide emissions,
climate change impacts removed
With worldwide GHG policy (01)
With worldwide GHG policy and mitigation (02)
With NZ-only ETS (06)
With NZ-only ETS and mitigation (07)
All
livestock
-1.4
-14.4
Beef
-9.2
-20.1
-7.0
-18.6
-6.9
-18.4
Sheepmeat Dairy
-1.7
4.0
-15.0
-9.9
-7.3
-19.2
-6.8
-18.0
The second two scenarios in Table 11 examine the impact of a New Zealand-only GHG
policy. For the first of the two, mitigation is not included. With the policy, emissions are
reduced from all the commodities. The reduction in emissions is increased by the addition of
mitigation policies. The reductions are fairly even across all the commodities.
The impact of consumer demand for low-emissions products was also considered in this set of
scenarios. Table 12 presents the results for three scenarios that include a price premium. The
first scenario has the price premium alone, which leads to a large increase in producer returns.
All four commodities have increased returns, but they have different levels of increases. A
second scenario includes Annex 1 GHG policies as well as mitigation technologies, and this
leads to even larger increases in producer returns. In the third scenario, only NZ implements a
GHG policy. All commodities still have increases in producer returns, although these
increases are lower than in the other two scenarios. For all scenarios, dairy has the largest
increase in returns, while forestry and beef have the lowest returns.
Table 12: Percentage changes in New Zealand producer returns with demand for
lower-emissions products
Demand impacts only (10)
Demand plus GHG policies and
mitigation (12)
Demand plus NZ-only ETS and
mitigation (17)
All
agriculture
31.6
Beef
15.1
47.1
18.6
40.8
74.9
10.7
25.3
8.7
24.9
40.6
10.6
Sheepmeat Dairy Roundwood
28.5
43.0
10.7
Table 13 presents the impact on GHG emissions from these same three scenarios. When
demand for low-emissions products is considered by itself, it leads to a general reduction in
GHG emissions. However, emissions from dairy increase as a result of increased production
to supply demand. In the scenario including both Annex 1 GHG policies and mitigation, there
is again a general decline in New Zealand emissions from livestock, but an increase in
emissions from dairy. In the final scenario, the New Zealand-only GHG policy and mitigation
18
lead to reduced emissions from a combination of the mitigation technologies and lowered
production.
Table 13: Percentage changes in New Zealand methane and nitrous oxide emissions,
with demand for lower-emissions products
Demand impacts only (10)
Demand plus GHG policies and mitigation
(12)
Demand plus NZ-only ETS and mitigation (17)
All
livestock
-6.2
Beef
-16.8
Sheepmeat
-6.0
Dairy
0.6
-7.3
-11.9
-23.5
-21.9
-7.6
-11.9
3.9
-5.3
Conclusion
This research created a model of international trade with unique capabilities. It extended the
LTEM, an existing model of trade in agricultural commodities that includes modules for
assessing GHG emission and energy, by incorporating data and parameters from the GFPM,
and international model of trade in forestry products. This extended LTEM allows the impact
of commodity price fluctuations on switching between agricultural and forestry production to
be endogenised in a single model, and the impact on prices, production, and GHG emission to
be assessed at the commodity and country levels.
The extended LTEM was then used to model 15 scenarios regarding the potential impacts of
climate change, GHG policies in Annex 1 countries and New Zealand, emissions mitigation
efforts in New Zealand, and consumer reactions to products produced with lower emissions.
Scenarios were developed that examine each of these items alone as well as in combinations.
These scenarios were modelled, and the impacts on producer returns and GHG emissions
were calculated for New Zealand agriculture and for the three livestock sectors, beef, sheep,
and dairy.
The results suggest that net impacts may be negative or positive for New Zealand, depending
on the actual effects of climate change, the policies enacted, and efforts at mitigation
emissions and linking emissions reductions to the market. In general, the results suggest five
tentative conclusions:
1. Climate change and worldwide GHG policies may improve returns for New Zealand’s
primary sector. These changes will reduce agricultural productivity abroad, increasing
worldwide prices and potentially increase demand for exports from New Zealand.
2. An ETS of itself may have a small impact on GHG emissions. The cost of carbon
credits is a small fraction of the total income from agriculture, so the reduction in
production is likely also to be small.
3. An ETS may be effective in reducing emissions if combined with support for
mitigation or marketing. The greatest benefit of an ETS may be in enabling and
promoting efforts to link emissions reductions to payments and premium markets.
4. Mitigation may be effective, and may benefit from government support. This research
modelled emissions mitigation that is experimentally possible but may not have been
achieved on farm. The impacts on total emissions were significantly larger than
without mitigation efforts.
19
5. Promoting New Zealand products as low-emissions products is likely to improve
producer returns in New Zealand. Achieving higher price for primary sector products
means higher producer returns, lower emissions, and greater productivity in the sector.
A large part of this research has been focused on developing the new model and working
through various issues that arose in combing the two models. These issues point to elements
of uncertainty regarding the modelling to date and the results presented, and indicate a
number of areas for future work.
•
Sensitivity analysis. With any new model, it is important to identify the key variables
or parameters that have large effects on the results, and then to determine the
sensitivity of the model to the initial values. In particular, it would be important to
investigate the equations and parameters that link the forestry products to the rest of
the model. There are also several parameters that affect some of the modelling results
presented here, such as the substitutability of different commodities. Some sensitivity
analysis of the model would allow researchers and policy-makers to understand the
areas where results are particularly robust, as well as areas where additional primary
research is needed to increase the certainty about model inputs.
•
Land use change. The original LTEM is a synthetic model, and relies on estimated
parameters to control the switching between commodities in production. Because of
the amount of work on this topic, it is possible to have relatively robust results. The
topic of land use, land use change, and forestry (LULUCF) is quite important when
discussing climate change and the carbon economy. Now that forestry is included in
the extended LTEM, it is possible to model LULUCF more directly, rather than using
the indirect approach.
•
Price of GHG emissions. The present research used one single price for GHG
emissions, US$25 per tonne of CO2-equivalent. There are a number of questions that
can now be investigated further. First, it would be interesting to investigate the impact
of the price level in combination with the other impacts included in the model.
Secondly, the impact of differential pricing by country, region, commodity, or
production method could be studied using the model, and may provide useful results
for understanding the potential impacts of different policies.
•
Mitigation. The technique for modelling mitigation in the present research
demonstrated the model capability, but it would be possible to develop more
sophisticated techniques. Such techniques could link the structure of production, land
use, policy, and markets.
•
Biofuels. The original LTEM has already been used to model biofuels, both
bioethanol and biodiesel, and the impact of biofuel policy on New Zealand agriculture.
With the earlier addition of sugar and now the extension to included forestry products,
it would be possible to investigate the impacts of biofuel policy that included several
different feedstocks. In addition, the linkage to the energy markets and GHG
emissions would allow a full investigate of the impact of new technologies and
policies on production, energy prices, and GHG emissions.
•
Different GHG policies. One of the core capabilities of the LTEM is modelling
domestic and international policies. For the present research, one type of policy – a
direct or indirect price on carbon – was considered. However, there are greater
capabilities in the model for investigating a number of policies, including
20
countervailing carbon tariffs, domestic subsidies for emissions reduction, and more.
By comparing these methods for reducing emissions, the impacts and unintended
consequences of policy can be investigated.
This research demonstrates both the difficulties and value of economic modelling for
understanding complex systems and the combinations and shocks and policies that affect
them. Although the two models, the GFPM and the LTEM, are both partial equilibrium trade
models, differences in specifications created considerably difficulty in incorporating the one
into the other. However, having done that, it was possible to model a number of scenarios and
estimate some initial results. The results, described above, will hopefully be useful in
designing New Zealand’s responses to climate change and international policy developments.
21
References
Buongiorno, J., C.S. Liu, and J. Turner. 2001. Estimating international wood and fiber
utilization accounts in the presence of measurement errors. Journal of Forest
Economics 7(2): 101-124.
Buongiorno, J., S. Zhu, D. Zhang, J.A. Turner, and D. Tomberlin. 2003. The Global Forest
Products Model: Structure, Estimation and Applications. Academic Press, San Diego.
301 pp.
Cesar, M. 1995. Global outlook for recovered paper. In Tappi Global Fiber Supply
Symposium, Tappi Press, Atlanta, GA. pp. 81-94.
Energy Information Administration (EIA). 2004. Energy Outlook Study 2004. Department of
Energy, Energy Information Administration, Washington D.C.
Food and Agriculture Organisation (FAO). 2005. Forest Resource Assessment 2005. Food
and Agriculture Organization of the United Nations, Rome.
www.fao.org/forestry/site/fra/en (Accessed 14 November 2005).
Food and Agriculture Organisation (FAO). 2008. Online FAO Yearbook of Forest Products,
FAOSTAT statistics database. Food and Agriculture Organization of the United
Nations, Rome. http://apps.fao.org/ (Accessed 16 May 2008)
Ince, P.J. 1994. Recycling and long-range timber outlook. General Technical Report GTRRM-242. USDA Forest Service. 23 pp. Mabee, W.E. 1998. The importance of
recovered fibres in global fibre supply. Unasylva 49(193): 31-36.
Kaye-Blake, W., Greenhalgh, S., Turner, J., Holbek, E., Sinclair, R., and Matunga, T. 2009. A
Review of Research on Economic Impacts of Climate Change. AERU Research
Report #314. Lincoln, NZ: Agribusiness and Economics Research Unit, April.
Mabee, W.E., and H. Pande. 1997. Recovered and non-wood fiber: Effects of alternative
fibers on global fiber supply. Working Paper GFSS/WP/04. Food and Agriculture
Organization of the United Nations.
Organization for Economic Cooperation and Development (OECD). 2004. OECD Economic
Outlook No. 76, January 2005. Organization for Economic Cooperation and
Development, Paris.
Cagatay, S., Saunders, C. M. and Wreford, A. 2003. Lincoln Trade and Environment Model
(LTEM): Linking Trade and Environment. AERU Research Report No. 263, Lincoln
University, NZ.
Cagatay, S.and Saunders, C. 2003. Lincoln Trade and Environment Model (LTEM): An
Agricultural Multi-Country, Multi-Commodity Partial Equilibrium Framework.
AERU Research Report No. 254, Lincoln University, NZ.
Simangunsong, B., and J. Buongiorno. 2001. International demand for forest products: A
comparison of methods. Scandinavian Journal of Forest Research 16: 155-172.
Turner, J., and J. Buongiorno. 2001. International freight rates for forest products: Structure,
past trends and forecasts. International Forestry Review 3(2): 136-145.
Turner, J.A., J. Buongiorno, S. Zhu, and R. Li. 2005. Calibrating and updating the Global
Forest Products Model (GFPM version 2005). Staff Paper Series #57. Department of
Forest Ecology and Management, University of Wisconsin-Madison. Madison, WI.
22
Turner, J.A., J. Buongiorno, and S. Zhu. 2006a. An economic model of international wood
supply, forest stock and forest area change. Scandinavian Journal of Forest Research
21: 73-86.
Turner, J.A., J. Buongiorno, F. Maplesden, S. Zhu, S. Bates, and R. Li. 2006b. World Wood
Industries Outlook: 2005-2030. Forest Research Bulletin 230. Ensis, Rotorua, New
Zealand. 84 p. ISBN 0-478-11014-6.
United Nations. 2004. World Population Prospects: The 2004 Revision. Population Division
of the Department of Economic and Social Affairs of the United Nations Secretariat.
http://esa.un.org/unpp, (Accessed 14 June 2005).
United Nations. 2005. World Economic Situation and Prospects 2005. Department of
Economic and Social Affairs, United Nations, New York.
United Nations Economic Commission for Europe/ Food and Agriculture Organization
(UNECE/FAO). 2002. Forecasts of the Economic Growth in OECD Countries and
Central and Eastern European Countries for the Period 2000-2040. United Nations,
New York.
World Bank. 2005. World Development Indicators online database. World Bank,
Washington, D.C.
23
Appendix
Table 1: Country codes in GFPM 12
Code Country
AFRICA
A0
Algeria
A1
Angola
A2
Benin
A3
Botswana
A4
Burkina Faso
A5
Burundi
A6
Cameroon
A7
Cape Verde
A8
Central African Republic
A9
Chad
B0
Congo, Republic of
B1
Côte d'Ivoire
B2
Djibouti
B3
Egypt
B4
Equatorial Guinea
B5
Ethiopia
B6
Gabon
B7
Gambia
B8
Ghana
B9
Guinea
C0
Guinea-Bissau
C1
Kenya
C2
Lesotho
C3
Liberia
C4
Libyan Arab Jamahiriya
C5
Madagascar
C6
Malawi
C7
Mali
C8
Mauritania
C9
Mauritius
D0
Morocco
D1
Mozambique
D2
Niger
D3
Nigeria
D4
Réunion
D5
Rwanda
D6
Sao Tome and Principe
D7
Senegal
D8
Sierra Leone
D9
Somalia
E0
South Africa
E1
Sudan
E2
Swaziland
E3
Tanzania, United Rep of
E4
Togo
E5
Tunisia
E6
Uganda
E7
Congo, Dem Republic of
E8
Zambia
E9
Zimbabwe
12
Code
Country
N/C AMERICA
Code
Country
ASIA
Code
Country
EUROPE
F0
F1
F2
F3
F4
F5
F6
F7
F8
F9
G0
G1
G2
G3
G4
G5
G6
G7
G8
G9
H0
H1
Bahamas
Barbados
Belize
Canada
Cayman Islands
Costa Rica
Cuba
Dominica
Dominican Republic
El Salvador
Guatemala
Haiti
Honduras
Jamaica
Martinique
Mexico
Netherlands Antilles
Nicaragua
Panama
Saint Vincent/Grenadines
Trinidad and Tobago
United States of America
SOUTH AMERICA
H2
H3
H4
H5
H6
H7
H8
H9
I0
I1
I2
I3
I4
Argentina
Bolivia
Brazil
Chile
Colombia
Ecuador
French Guiana
Guyana
Paraguay
Peru
Suriname
Uruguay
Venezuela, Boliv Rep of
I5
I6
I7
I8
I9
J0
J1
J2
J3
J4
J5
J6
J7
J8
J9
K0
K1
K2
K3
K4
K5
K6
K7
K8
K9
L0
L1
L2
L3
L4
L5
L6
L7
L8
L9
M0
M1
M2
M3
Afghanistan
Bahrain
Bangladesh
Bhutan
Brunei Darussalam
Cambodia
China
Cyprus
Hong Kong
India
Indonesia
Iran, Islamic Rep of
Iraq
Israel
Japan
Jordan
Korea, Dem People's Rep
Korea, Republic of
Kuwait
Laos
Lebanon
Macau
Malaysia
Mongolia
Myanmar
Nepal
Oman
Pakistan
Philippines
Qatar
Saudi Arabia
Singapore
Sri Lanka
Syrian Arab Republic
Thailand
Turkey
United Arab Emirates
Viet Nam
Yemen
OCEANIA
N5
N6
N7
N8
N9
O0
O1
O2
O3
O4
O5
O6
O7
O8
O9
P0
P1
P2
P3
P4
P5
P6
P7
P8
P9
Q0
Q1
Q2
Q3
Q4
Albania
Austria
Belgium
Bosnia and Herzegovina
Bulgaria
Croatia
Czech Republic
Denmark
Finland
France
Germany
Greece
Hungary
Iceland
Ireland
Italy
Macedonia, The Fmr Yug Rp
Malta
Netherlands
Norway
Poland
Portugal
Romania
Slovakia
Slovenia
Spain
Sweden
Switzerland
United Kingdom
Serbia and Montenegro
FORMER USSR
M4
M5
M6
M7
M8
M9
N0
N1
N2
N3
N4
Australia
Cook Islands
Fiji Islands
French Polynesia
New Caledonia
New Zealand
Papua New Guinea
Samoa
Solomon Islands
Tonga
Vanuatu
Q5
Q6
Q7
Q8
Q9
R0
R1
R2
R3
R4
R5
R6
R7
R8
R9
Armenia
Azerbaijan, Republic of
Belarus
Estonia
Georgia
Kazakhstan
Kyrgyzstan
Latvia
Lithuania
Moldova, Republic of
Russian Federation
Tajikistan
Turkmenistan
Ukraine
Uzbekistan
ZY
ZZ
Dummy Region
World
The listed countries are default countries in GFPM. To add or remove countries, see Zhu et al. (2008). 24
Table 2: Wood products in the Global Forest Products Model (GFPM)
Commodity Aggregate
Constituent Commodities
(used in the GFPM)
Fuelwood and charcoal
Wood fuel
Wood charcoal
Industrial roundwood
Chips and particles (imports and exports only)
Pulpwood
Sawlogs
Other industrial roundwood
Other industrial roundwood
Sawnwood
Sawnwood
Plywood
Plywood
Veneer sheets
Particleboard
Particleboard
Fibreboard
Fibreboard
Mechanical pulp
Mechanical wood pulp
Chemical pulp
Chemical wood pulp
Semi-chemical wood pulp
Other fibre pulp
Other fibre pulp
Waste paper
Recovered paper
Newsprint
Newsprint
Printing and writing paper
Printing and writing paper
Other paper and paperboard
Other paper and paperboard
25
Table 3: Price and income elasticities of demand for final products
Commodity
Fuelwood
Other industrial roundwood
Sawnwood
Plywood and veneer
Particleboard
Fibreboard
Newsprint
Printing and writing paper
Other paper and paperboard
Wealth - Region
Price
Income
High income1
-0.62
-1.50
Low income2 – Africa
-0.10
0.40
Low income – Other regions
-0.10
0.15
High income
-0.05
-0.58
Low income
-0.37
0.19
High income
-0.16
0.32
Low income
-0.21
0.46
High income
-0.13
0.10
Low income – Europe
-0.22
1.20
Low income – Other regions
-0.22
0.74
High income
-0.24
1.25
Low income
-0.05
0.65
High income
-0.52
0.82
Low income – Asia, Europe
-0.52
1.50
Low income – Other regions
-0.52
1.10
High income
-0.05
0.21
Low income – Asia, Europe
-0.18
1.05
Low income
-0.18
0.21
High income
-0.15
0.80
Low income
-0.37
1.11
High income
-0.06
0.65
Low income
-0.14
0.92
1
Australia, Austria, Belgium-Luxembourg, Canada, Denmark, Finland, France, Germany, Ireland, Israel, Italy,
Japan, Kuwait, Netherlands, Norway, New Zealand, South Africa, Spain, Sweden, Switzerland, United
Kingdom, and the USA
2
Rest of the world
Modified from Buongiorno et al. (2003, Table 4.5).
26
Table 4: Equation parameters for fuelwood and industrial roundwood supply in the
Global Forest Products Model.
Fuelwood Supply
Price
Industrial Roundwood Supply
Low income
High income
Low income
1.00
2.00
0.40-1.571
0.70-1.57
0.90
0.90
1.60
0.50
GDP per capita
Forest stock
1.00
1.50
High income
1
The price elasticity of industrial roundwood supply depends on the proportion of country forest in public
ownership; with supply from public forests less price elastic than supply from private forests.
27
Table 5: Supply shift parameters for GHG policy scenarios
Commodity
Wheat
Other grains
Maize
Rice
Sugar
Oilseed
Oilseed meal
Oil
Beef
Pork
Sheepmeat
Wool
Poultry
Eggs
Raw milk
Liquid milk
Butter
Cheese
Whole milk powder
Skim milk powder
Apples
Kiwifruit
Firewood
Roundwood
Panelwood
Paper and pulp
Australia Canada
1.00
0.96
1.00
0.96
1.00
0.96
1.00
0.96
1.00
0.96
1.00
0.96
1.00
0.96
1.00
0.96
0.74
0.87
1.00
0.96
0.78
0.86
0.78
0.86
1.00
0.96
1.00
0.96
0.88
0.93
0.88
0.93
0.88
0.93
0.88
0.93
0.88
0.93
0.88
0.93
1.00
0.96
1.00
0.96
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
European
Union
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.92
0.95
0.87
0.87
0.95
0.95
0.96
0.96
0.96
0.96
0.96
0.96
0.95
0.95
1.00
1.00
1.00
1.00
Japan
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.87
0.96
0.86
0.86
0.96
0.96
0.93
0.93
0.93
0.93
0.93
0.93
0.96
0.96
1.00
1.00
1.00
1.00
28
New
Zealand
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.93
0.94
0.91
0.91
0.94
0.94
0.92
0.92
0.92
0.92
0.92
0.92
0.94
0.94
1.00
1.00
1.00
1.00
Norway
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.87
0.96
0.86
0.86
0.96
0.96
0.93
0.93
0.93
0.93
0.93
0.93
0.96
0.96
1.00
1.00
1.00
1.00
Russia
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.87
0.96
0.86
0.86
0.96
0.96
0.93
0.93
0.93
0.93
0.93
0.93
0.96
0.96
1.00
1.00
1.00
1.00
Switzerland
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.87
0.96
0.86
0.86
0.96
0.96
0.93
0.93
0.93
0.93
0.93
0.93
0.96
0.96
1.00
1.00
1.00
1.00
Turkey
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.87
0.96
0.86
0.86
0.96
0.96
0.93
0.93
0.93
0.93
0.93
0.93
0.96
0.96
1.00
1.00
1.00
1.00
USA
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.94
0.89
0.94
0.90
0.90
0.94
0.94
0.96
0.96
0.96
0.96
0.96
0.96
0.94
0.94
1.00
1.00
1.00
1.00
Table 6: Supply shift parameters for mitigation scenarios
Commodity
Wheat
Other grains
Maize
Rice
Sugar
Oilseed
Oilseed meal
Oil
Beef
Pork
Sheepmeat
Wool
Poultry
Eggs
Raw milk
Liquid milk
Butter
Cheese
Whole milk powder
Skim milk powder
Apples
Kiwifruit
Firewood
Roundwood
Panelwood
Paper and pulp
Australia Canada
1.00
0.97
1.00
0.97
1.00
0.97
1.00
0.97
1.00
0.97
1.00
0.97
1.00
0.97
1.00
0.97
0.82
0.91
1.00
0.97
0.84
0.90
0.84
0.90
1.00
0.97
1.00
0.97
0.91
0.95
0.91
0.95
0.91
0.95
0.91
0.95
0.91
0.95
0.91
0.95
1.00
0.97
1.00
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
European
Union
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.94
0.97
0.91
0.91
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
1.00
1.00
1.00
1.00
Japan
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.91
0.97
0.90
0.90
0.97
0.97
0.95
0.95
0.95
0.95
0.95
0.95
0.97
0.97
1.00
1.00
1.00
1.00
29
New
Zealand
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.95
0.96
0.94
0.94
0.96
0.96
0.94
0.94
0.94
0.94
0.94
0.94
0.96
0.96
1.00
1.00
1.00
1.00
Norway
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.91
0.97
0.90
0.90
0.97
0.97
0.95
0.95
0.95
0.95
0.95
0.95
0.97
0.97
1.00
1.00
1.00
1.00
Russia
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.91
0.97
0.90
0.90
0.97
0.97
0.95
0.95
0.95
0.95
0.95
0.95
0.97
0.97
1.00
1.00
1.00
1.00
Switzerland
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.91
0.97
0.90
0.90
0.97
0.97
0.95
0.95
0.95
0.95
0.95
0.95
0.97
0.97
1.00
1.00
1.00
1.00
Turkey
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.97
0.91
0.97
0.90
0.90
0.97
0.97
0.95
0.95
0.95
0.95
0.95
0.95
0.97
0.97
1.00
1.00
1.00
1.00
USA
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.92
0.96
0.93
0.93
0.96
0.96
0.97
0.97
0.97
0.97
0.97
0.97
0.96
0.96
1.00
1.00
1.00
1.00
Table 7: Supply shift parameters for climate change scenarios
Commodity
Argentina
Australia
Brazil
Canada
China
EU
India
Japan
Korea
Mexico
New
Zealand
Norway
Russia
South
Africa
Switzerland
Turkey
USA
ROW
Wheat
Other grains
Maize
Rice
Sugar
Oilseed
Oilseed meal
Oil
Beef
Pork
Sheepmeat
Wool
Poultry
Eggs
Raw milk
Liquid milk
Butter
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.93
0.96
0.93
0.93
0.96
0.96
0.96
0.96
0.96
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.93
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
0.96
Cheese
Whole milk powder
Skim milk powder
Apples
Kiwifruit
Firewood
Roundwood
Panelwood
Paper and pulp
0.96
0.96
0.96
0.96
0.96
1.01
1.01
1.01
1.01
0.99
0.99
0.99
0.99
0.99
1.01
1.01
1.01
1.01
0.96
0.96
0.96
0.96
0.96
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
0.93
0.93
0.93
0.93
0.93
1.01
1.01
1.01
1.01
0.99
0.99
0.99
0.99
0.99
1.01
1.01
1.01
1.01
0.93
0.93
0.93
0.93
0.93
1.01
1.01
1.01
1.01
0.99
0.99
0.99
0.99
0.99
1.02
1.02
1.02
1.02
0.99
0.99
0.99
0.99
0.99
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
0.96
0.96
0.96
0.96
0.96
1.11
1.11
1.11
1.11
0.99
0.99
0.99
0.99
0.99
1.02
1.02
1.02
1.02
0.93
0.93
0.93
0.93
0.93
1.01
1.01
1.01
1.01
0.96
0.96
0.96
0.96
0.96
1.01
1.01
1.01
1.01
0.99
0.99
0.99
0.99
0.99
1.01
1.01
1.01
1.01
0.99
0.99
0.99
0.99
0.99
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.02
1.02
1.02
1.02
0.96
0.96
0.96
0.96
0.96
1.01
1.01
1.01
1.01
30